Model profile
Qwen3.5 35B A3B
Evidence summary
Qwen3.5 35B A3B has an estimated overall rank of #175; its 90% source-sensitivity interval is #39–#255. Its behavior-only rank is #172; company governance moves the combined estimate to #175. Published evidence spans 3 evals and 4 of 7 behavior components. Its strongest relative result is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #21 of 260); its weakest is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #238 of 260).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗qwen3-5-35b-a3b
- Hugging Face ↗Qwen/Qwen3.5-35B-A3B
- OpenRouter ↗qwen/qwen3.5-35b-a3b
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗repository created at proxy
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Better | Source |
|---|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #187 / 311 | 0.8401 | ↓ lower | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #57 / 260 | 26.61 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #238 / 260 | 73.17 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #21 / 260 | 98.33 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #62 / 258 | 98.36 | ↑ higher | Source ↗official | |
| SABERoverall_safety_rate | #10 / 13 | 22.73 | ↑ higher | Source ↗official | |
| SABERscenario_a_safety_rate | #11 / 13 | 23.55 | ↑ higher | Source ↗official | |
| SABERscenario_b_safety_rate | #7 / 13 | 32.73 | ↑ higher | Source ↗official | |
| SABERscenario_c_safety_rate | #8 / 13 | 14.54 | ↑ higher | Source ↗official |
Values evaluations
Descriptive values and political-framing results are separate from safety/ethics ranks. Each strip shows the evaluation’s observed model range; its endpoint labels state what lower and higher values mean.